Genomic imprinting results in the specific expression of parental alleles for endogenous genes that play significant roles in growth regulation of mammalian embryos. Disregulation of these genes appears to be associated with neoplasias and neonatal cancers in humans and a number of developmental syndromes. Five imprinted genes have been characterized in the mouse but the total number of imprinted genes and their potential roles in cancer and other diseases is unknown. A method of screening the mouse genome for new imprinted genes has been developed that is based upon differential methylation of CpG islands. This proposal describes restriction landmark genomic scanning (RLGS) which uses high resolution two-dimensional gel electrophoresis to identify end-labeled, rare-site cleavages of genomic DNA using Not I and similar endonucleases. Two thousand RLGS landmarks can be identified as spots in a single two dimensional gel and additional landmarks for a genome can be readily produced with different restriction enzyme combinations. Genetic variation for these landmarks has been identified between common inbred strains and between laboratory mice and other Mus species. We have screened more than 2000 strain-specific sites in the mouse genome for differential methylation of parental genes using a landmark enzyme that is methylation sensitive (NotI). Parent allele-specific NotI digestion was detected by analyzing reciprocal F1 combinations between various mouse genotypes. Eight loci were identified that showed differential methylation of parental genes. One locus on mouse chromosome 11 has been characterized which shows a derived amino acid sequence homology with the U2 small nucleoprotein auxiliary factor small subunit -U2AF binding protein. Molecular clones for a second locus have been identified and work is in progress to characterize the imprinted gene. This proposal outlines an experimental program that will characterize five additional imprinted RLGS loci (Irlgs loci) that have been identified as a first step in determining the molecular basis for imprinting and the evolutionary reasons for sustaining monoallelic expression of different genes. We also propose to screen an additional 3500 - 4000 landmarks in the mouse genome for new Irlgs loci as a means of determining the total number of endogenous genes that are imprinted. The cloned sequences for imprinted mouse genes can be used to identify the homologous human genes and their linkage in the human genome. This information will provide a more complete opportunity to establish the number of imprinted genes in the mammalian genome and their potential involvement in cancer and other developmental disorders in humans.